Monitoring ND-PM Communication Health
The ND-PM Communication Observability Monitoring provides real-time visibility into the communication health between Network Director (ND) and Policy Manager (PM) containers, including channel status, health conditions, and performance metrics.
On this page:
- Enable ND-PM Communication Monitoring on an ND Container by using a recipe
- Interpret the ND-PM communication health metrics
Enable ND-PM Communication Monitoring on an ND Container by using a recipe
You can enable the ND-PM Communication Observability Monitoring on each ND container by using a recipe to monitor ND-PM communication health metrics.
Use the following recipe to install the ND-PM communication monitoring feature. Run the recipe by using the following structure:
./jython.sh -m akana.container --recipe ../recipes/install-nd-pm-communication-monitoring-feature.json --props <ND_PROPERTIES_FILE>
The following shows the content of the install-nd-pm-communication-monitoring-feature.json recipe:
{
"name":"install-nd-pm-communication-monitoring-feature",
"description":"Install ND-PM communication observability on ND container (pmcomm base + pmcomm.nd). Runs post-registration so container-identity task is already complete.",
"phases":["deploy"],
"bundles":{
"install":[
{"location":"obr://com.akana.observability.pmcomm"},
{"location":"obr://com.akana.observability.pmcomm.nd"}
]
},
"configurations":[
{
"pid":"com.akana.observability.pmcomm.nd",
"update":{
"pmcomm.channel.config.calculation.window.ms":"30000",
"pmcomm.channel.engine.calculation.window.ms":"30000",
"pmcomm.channel.usage.calculation.window.ms":"60000",
"pmcomm.channel.registration.calculation.window.ms":"60000",
"pmcomm.channel.auth.calculation.window.ms":"300000",
"pmcomm.channel.alerts.calculation.window.ms":"600000"
}
}
]
}
Interpret the ND-PM communication health metrics
The ND-PM communication monitoring provides detailed information about the health and performance of communication channels between the Network Director and Policy Manager containers.
Communication channels
The ND-PM communication is organized into the following channels. Each channel monitors specific types of requests between the Network Director and Policy Manager, with defined expected cadences and tracked URLs.
| Channel | Purpose | Expected Cadence | Tracked URLs |
|---|---|---|---|
| config | Configuration queries and container configuration requests. | 30000 ms |
|
| engine | Policy engine calls and policy evaluation requests. | 30000 ms |
|
| usage | Usage monitoring and metrics publication. | 60000 ms |
|
| registration | Container registration requests. | 60000 ms |
|
| auth | Authentication and authorization requests. | 300000 ms |
|
| alerts | Alert management and reporting requests. | 600000 ms |
|
Health statuses
A communication channel’s health status is determined by the recency of successful and failed checks.
| Status | Description |
|---|---|
| HEALTHY | Indicates whether the most recent check succeeded or failed. |
| DEGRADED | Indicates that both a recent failure and a recent success were observed. |
| DOWN | Indicates a recent failure with no successful check within the calculation window. |
Health rollup rule
The pm-nd.health status is calculated from the six per-channel verdicts. The rules are evaluated in the following order:
| Condition | Overall pm-nd.health rollup |
|---|---|
| At least one channel has a DOWN verdict | DOWN |
| No channel is DOWN, and at least one channel has a DEGRADED verdict | DEGRADED |
| All channels have a HEALTHY verdict | HEALTHY |
| A channel has an UNKNOWN verdict | Excluded from the calculation. It does not lower the overall health status. |